@inproceedings{oai:ynu.repo.nii.ac.jp:00010078,
 author = {Kinjo, Yukihiro and Sakuma, Yoshitomo and Kohno, Ryuji},
 book = {IEICE Technical Report},
 issue = {24},
 month = {May},
 note = {In rehabilitation, approach according to personality is important.
So we estimate patients' emotion by neural network(NN) for their R-R interval(RRI) in heart rate data from Wireless Body Area Network(WBAN). However, machine learning processing is complexity and sending heart rate data to super computer like Watson for machine learning processing causes network delay. In this research, we propose how to reduce computational complexity to enable to calculate by limited processing power. Specifically, we aim to reduce it to use NN with preprocessing by wavelet transform and extraction of coefficient of variance of RRI(CVRR). Preprocessing extract a part of characteristic before processing of NN and computational complexity by NN processing reduce.},
 pages = {101--104},
 publisher = {IEICE},
 title = {Learning and recognition with neural network of heart beats sensed by WBAN for stress estimate for rehabilitation},
 volume = {118},
 year = {2018}
}